Two undergraduate researchers from the Data-Driven Decision Making (D3M) Lab successfully completed their projects through Penn State’s Multi-Campus Research Experience for Undergraduates (MCREU) program, culminating in research presentations, technical reports, and publication of their project summaries on the MCREU website.
The projects span two of the D3M Lab’s core research areas: Biomedical Image Segmentation and AI for Mental Health Analytics, demonstrating the lab’s commitment to providing undergraduate students with hands-on interdisciplinary research experiences.
Project 1: Automated Multi-Class Segmentation of Zebrafish Micro-CT Scans Using a Deep Learning U-Net
Under the mentorship of Dr. Sayed Mohsin Reza and collaborators from Penn State College of Medicine, undergraduate researcher Dhruv Sanghavi developed a deep learning framework for automated segmentation of zebrafish micro-CT images.
The project explored the use of a ResNet-34 U-Net architecture to automate the identification and segmentation of anatomical structures, including the brain, heart, and eyes. Automated biomedical image segmentation has the potential to significantly accelerate the creation of anatomical atlases and reduce the time required for manual expert annotation.
The results demonstrated strong segmentation performance and highlight the potential of deep learning techniques to support large-scale biomedical imaging research. The work contributes to ongoing D3M Lab efforts in developing AI-driven tools for biomedical image analysis and digital biological atlases.
Student Researcher: Dhruv Sanghavi
Faculty Mentors: Dr. Sayed Mohsin Reza and Dr. Khai Chung Ang
Project 2: Interpretable Machine Learning for Clinical Risk Detection in Personality Disorders
Undergraduate researcher Om Patel conducted research focused on applying interpretable machine learning approaches to personality disorder data in order to identify clinically meaningful risk factors associated with mental health outcomes.
Working with Dr. Sayed Mohsin Reza and Dr. Sharon Nelson, the project investigated whether personality assessment data could be used to predict elevated suicide ideation risk, treatment rejection, and potentially under-recognized personality pathology. The research utilized interpretable ensemble machine learning models combined with explainable AI techniques to ensure that model predictions remained transparent and clinically meaningful.
The project demonstrated the potential of machine learning to assist in identifying at-risk individuals while providing explainable insights that may support future clinical decision-making and mental health research.
Student Researcher: Om Patel
Faculty Mentors: Dr. Sayed Mohsin Reza and Dr. Sharon Nelson
Advancing Undergraduate Research Through D3M Lab
The successful completion of these MCREU projects reflects the D3M Lab’s mission of integrating artificial intelligence, machine learning, biomedical research, and data-driven decision-making while providing meaningful research opportunities for undergraduate students.
Through the MCREU program, students gained experience in research design, scientific communication, literature review, poster development, and presentation of findings to the broader research community. These projects contribute to the growing portfolio of D3M Lab research in healthcare AI, biomedical imaging, and predictive analytics.
The D3M Lab congratulates Dhruv Sanghavi and Om Patel on their outstanding accomplishments and thanks the Penn State MCREU program for its continued support of undergraduate research excellence.